Re: [CAD_CAM_EDM_DRO] retrofitting an older CNC

no falloff wrote:

>After using CNC mills both at a friends shop, and as school, I decided I just had to have a machine of my own. After saving for the better part of a year, I just bought a CNC Shizuoka mill with a Dynapath Delta 20 controller (pics at: www.cyclegarage.net/mill). I’m not sure if the controller is any good. It seems to work but I’m getting a fault on the Z axis.
>

Does it give any more detail on what the fault is? Is the Z servo amp
tripping off?
You might check the limit switches to make sure no Z axis switch is
sticking or
has a bad contact. Does the Z axis move any before it faults? Are there
any LEDs
on the servo amps to indicate status?

> I bought the machine figuring that the controler wasn’t working, and had planed to replace it (of course only knowing enough to be a danger to myself and others :-) ). It seems retrofit controller systems by Centroid, and Anilam etc. are way too expensive for my current budget. I was hoping to use as much of the existing hardware (servos, servo amps and electronics) as possible, and just use a PC as the controller. The mill is currently fitted with Baldor permanent magnet servos. The book for the servo drives (the only documentation I have) mentions that the drives are looking for PWM signals from the controller. I know that
> one could probably write a book in response to my questions, but I just need a general bit of clarification.
>
>
>
>1. Systems that use gecko drives don’t have absolute feed back? That is, the computer sends motion signals (step and direction) to the drive but has no way of knowing if the servo is in the position it is supposed to be in? How does this compare to the positional accuracy as compared to the current system?
>
> In general, that is true. I have an interface that allows the feedback to go all
>

the way back to the computer. I am working on an adaptor board to go
between Gecko
320 servo drivers and my board to make the connection simpler.

>
>2. Does the fact that mach 1 uses step and direction commands mean that it can’t be used with the existing servo boards in this machine? If not what controller software (EMC?) would work?
>
>

Yes (I think). But, Mach 2 is designed to be able to use servos in a
closed loop mode.
I am collecting the software needed to write a Mach 2 driver for my
boards. They do
work with EMC now.

>3. How do I interface it with the existing hardware? If the servo amps don’t except step and direction signals, then I assume that a simple breakout board such as those used with gecko systems won’t work. Likewise anyone have suggestions as to what will?
>
>

Well, first, I think you need to find out what the Z axis problem is. If
it is a bad servo
amp, spending a lot of time and money interfacing to them is not a good
plan.
Yes, a breakout board would be useless. Do you have any manuals for it?
What make are the servo amps? Can you get manuals or specs on them?
If you can get me specs on the PWM characteristics, I can see about a
special
version of my stepper controller board to do PWM. I think it would be
pretty easy to convert the step rate generators to PWM generators.

>4. These servos are rated a 120v and 10A (continuous) 39A (peak). It is my understanding that these can be used with gecko 320 drives, but will not achieve their maximum rated torque or reach their 2500RPM max rated speed. Is this correct? Will the tachometer need to be replaced with encoders in order work with this system? Rutex drives seem a better choice, true?
>
>

It already has encoders of some form, I'm pretty sure. They may be
linear or on the
motors or leadscrews. But, I'm pretty sure Centroid used encoders, not
resolvers.
Yes, Gecko 320 drivers won't get you full performance. Rutex has a
driver that
will do better.

>Lastly. How hard would it be to include the stepper motor controlled spindle speed, and tach into the new controller? Could this be done by defining it as say a 4th axis?
>
>

Yuchhh! It can be done, but it is a pretty ghastly way to do things. Does it
actually grind the stepper in response to S1234 spindle speed commands, until the
spindle speed is measured to be at the commanded RPM, or is there
just a faster and slower button? I guess if you have 3-phase power, you
don't NEED a VFD for the 3-phase conversion, but they are a pretty nice
way to do variable speed.

Jon